Bi-level iso-surface compression

Abstract

Methods, structures and systems for encoding and decoding isosurface data. An encoder process takes volume data and an isolevel as input and produces compressed isosurface data as output. The compressed isosurface data produced by an encoder process is composed of an occupancy image record, an optional intersection points record, and an optional normal vectors record. An occupancy image is compressed with a context-based arithmetic encoder. Compressed isosurface data can be stored in a data storage device or transmitted through a communication medium to a remote computer system, where the decoder process is executed. The decoder processes take compressed surface data as input and produce surface data as output. The decoder processes first reconstructs the occupancy image by decoding the occupancy image record. An in-core isosurface decoder process produces a polygon mesh as a surface representation. An out-of-core isosurface decoder process produces a set of oriented points as a surface representation.

Description

PRIORITY[0001]This application claims priority to a Provisional Application, having the same title, assigned, Ser. No. 60 / 400,202, and filed Jul. 31, 2002, which is incorporated herein by reference in its entirety for all purposes.FIELD OF THE INVENTION[0002]This invention relates to the fields of computer graphics and scientific visualization. It is more particularly related to representing isosurface data in compressed form, especially suitable for compact storage and fast transmission, and remote visualization.BACKGROUND OF THE INVENTION[0003]Polygon meshes are widely used in computer aided geometric design, geometric modeling, medical imaging, and computer graphics to represent surfaces in digital form. Polygon meshes are described in detail in U.S. Pat. No. 5,506,947 “Curve and Surface Smoothing Without Shrinkage,” by G. Taubin, and in the paper “A Signal Processing Approach To Fair Surface Design,” by G. Taubin, Siggraph'95 Conference Proceedings, August 1995, pages 351-358, b...

AI Technical Summary

Benefits of technology

[0031]Thus, this invention provides methods, apparatus and systems for a new and simple algorithm to compress isosurface data. This is the data extracted by isosurface algorithms from scalar functions defined on volume grids, and used to generate polygon meshes or alternative representations. Isosurfaces are in widespread use in medical imaging, and scientific computation. The main features of this invention are its extreme simplicity and high compression ratios, which are 5 to 25 times better than those obtained with general purpose mesh compression schemes.

[0033]An in-core isosurface decoder process produces a polygon mesh as a surface representation. This decoder process takes as input a compressed isosurface data with neither the optional intersection points record nor the normal vectors record. In this decoder the normalized intersection points are set to the default value one half, the normal vectors are given default values as a function of the neighboring intersection point values. A subsequent smoothing algorithm is used as a global predictor to improve the quality of the reconstructed polygon mesh.

Problems solved by technology

When the geometry information (vertex coordinates, and optionally normals, colors, and texture coordinates) is also taken into account, the cost per vertex increases considerably.

In addition, all of these approaches are incompatible with the out-of-core nature of isosurface extraction algorithms that visit the voxels in scan order.

In both cases the transmission time constitutes a major bottleneck because of the large file sizes involved.

In the first case, in addition to the size of the transmitted data, the burden of the computation is shifted to the client.

For binary data, where the alphabet is composed of two symbols, keeping track of global symbol frequencies is usually not good enough as a model update procedure, and a “context-based” procedure is used.

Images